Re: Adding normalise to gnuplot

[Jon G. <jo...@th...>]

> > One feature I've been missing from gnuplot is the ability to normalise a
> > dataset; that is, scale all values so that they are in [0:1] by dividing
> > by the maximum value. The attached patch does so by introducing "smooth
> > normal".
> 
> The usual thing to do to "normalise" data would be subtract the mean and 
> divide by std dev.

The reason I want to divide by the max here is that I want to ensure the
plot will always fit within the range I specify. "Proper" normalization
wouldn't do that. In particular, this is useful for live plots where the
shape of the data may change drastically from one plot to the next.

> I think the design philosophy here has been that most data processing is 
> well catered for elsewhere and GP is not trying to be kitchensink, one 
> stop , DP and plotting tool.

I agree. I do feel that some commonly used functions should be included
though, to prevent the need for external tools for common tasks.
I admit normalization (or variations thereof) might not fall into that
category though.

> However, if you want to do this sort of thing it does not require a 
> patch to gnuplot, just use a function.
> 
> max(x) =(if (x>biggest)?biggest=max:(x))
> biggest=-1e-134; plot datafile 1:max($2)
> plot datafile 1:($2/biggest)

Ah, of course, why didn't I think of that?!
Thanks, I'll use this instead.

Jon